[AlemayehuA]

Hi,

I need to define and solve a modified transport equation of additional variable (AV). The equation obtained from volume averaging of the transport equations (PDEs) in the gas phase and in solid phase of the porous domain is as shown below

a(dC/dt) + del(UC) = del(D)del(C) + R

where, a is a constant.

[ghorrocks]

You mean you want to add the "a" term in front of the transient term, but otherwise a normal transport equation? Why would you want to do this? What physical process does this model?

I would use a normal transport equation and put the modification to the transient term in a source term.

[AlemayehuA]

Originally Posted by ghorrocks

You mean you want to add the "a" term in front of the transient term, but otherwise a normal transport equation? Why would you want to do this? What physical process does this model?

I would use a normal transport equation and put the modification to the transient term in a source term.

Dear ghorrocks, thanks for your interest. Ya, its a normal transport equation. Its to model gas diffusion and adsorption in porous media. The equation arises after adding and simplifying the transport equations for the fluid phase and for the solid phase of the porous media.

What do you mean by putting the modification to the transient term in a source term? how do you do it? what happen to the convective and diffusive terms?

[ghorrocks]

Recast your equation in the form

(dC/dt) + del(UC) = del(D)del(C) + (R-(a-1)(dC/dt))

And you can now use the normal transport equation AV, and the term (R-(a-1)(dC/dt)) becomes your source term.

[AlemayehuA]

Thanks so much! this looks promising...then I plan to approximate the dC/dt term in the source as CEL expression given as C.Time Derivative, what do you think?

[AlemayehuA]

Look the error message I got,

Error detected by routine PSHDIR
CDRNAM = TIMEDERIV
CRESLT = NONE

may be not proper to try to simultaneously evaluate both the variable and its derivative

[ghorrocks]

No, I do not think CEL can access time derivatives. So you need to work out how to access time derivatives. There has been some posts about this on the forum, do a search for them.

[katty17]

Hi Glenn,


I want to model flow trough a series of connected porous walls. I need CFX solves the volume-averaged momentum and continuity equations. How can I ensure that the governing equations in CFX are solving in their volume-averaged form?


Many thanks in advance,
Katty

[ghorrocks]

CFX solves the Navier Stokes equations, or if turbulent the Reynolds Averaged (or Fauve Averaged) Navier Stokes equations. It does it by discretising the equations over control volumes, so each control volume can be considered volume averaged. Is that what you mean?

[katty17]

Hi Glenn,


Thanks for your response. Actually I was confused by the what "volume averaged" means in the porous domains. It seems it is exactly similar to the simple finite volume approach and just considers an additional source term to account for the porosity effect.